On Wed, Oct 7, 2020 at 12:10 PM Michal Simek <michal.simek@xxxxxxxxxx> wrote:
> On 07. 10. 20 10:55, Andy Shevchenko wrote:
> > On Wed, Oct 7, 2020 at 11:38 AM Michal Simek <michal.simek@xxxxxxxxxx> wrote:
> >> On 06. 10. 20 17:55, muhammad.husaini.zulkifli@xxxxxxxxx wrote:
...
> >>> + if (of_device_is_compatible(np, "intel,keembay-sdhci-5.1-sd")) {
> >>> + struct gpio_desc *uhs;
> >>> +
> >>> + uhs = devm_gpiod_get_optional(dev, "uhs", GPIOD_OUT_HIGH);
> >>
> >> I can't see change in dt binding to record uhs gpio.
> >>
> >>
> >> Better
> >> sdhci_arasan->uhs_gpio = devm_gpiod_get_optional(dev, "uhs",
> >> GPIOD_OUT_HIGH);
> >>
> >> then you can avoid uhs variable.
> >
> > Actually it's readability vs. additional variable. It was my
> > suggestion to have a variable to make readability better.
> > Are you insisting on this change?
>
> I understand that it is just about preference.
Correct.
> I would use it directly
> not to deal with it. If your preference is via variable I am fine with
> it too.
Yes, I suggested that and I still have the same opinion (It seems the
original code tries to follow 80 character rules that's why I
suggested the change).
Thanks!
--
With Best Regards,
Andy Shevchenko
